MathWorks joins forces with NVIDIA to accelerate the development of software-defined workflows in medical technology

Source: MathWorks Author: MathWorks

    Beijing, China, May 8, 2024 - MathWorks, the world‘s leading developer of mathematical computing software, today announced the successful integration of MATLAB into the NVIDIA Holoscan platform. Medical device engineers can now wrap existing MATLAB algorithms and functions into GPU-accelerated NVIDIA Holoscan operators for real-time data processing and inference, accelerating the development and deployment of streaming data analysis and visualization applications.

    Medical device engineers must not only adapt to complex and evolving global regulations, but also keep up with rapid technological innovation using cutting-edge materials and electronic components. These conflicting factors have led to many devices approaching obsolescence shortly after they hit the market, giving rise to medical device software (SaMD). The International Forum of Medical Device Regulators defines SaMD as software that is independent of a hardware medical device and capable of achieving one or more medical purposes. To remain competitive in the market after deployment, SaMD products need to be able to integrate additional software capabilities over time, which requires engineers to develop software-defined workflows to support development activities.

    David Niewolny, Director of Medical Technology Business Development at NVIDIA, said: "Artificial intelligence is transforming the medical technology industry. "The collaboration between NVIDIA and MathWorks provides a seamless development experience for the growing MATLAB development community within NVIDIA‘s medical-grade Holoscan platform, accelerating AI innovation in medical technology."

    NVIDIA Holoscan is a sensor processing platform that simplifies the development and deployment of AI and high-performance computing applications for real-time depth information. The platform provides a full-stack infrastructure for software-defined, scalable streaming data processing on edge devices, enabling the latest AI applications in clinical scenarios. With the integration of MATLAB and Holoscan, medical device engineers will be able to implement image and signal processing, filtering, transformation, and deep learning algorithms using existing built-in matrix operations and complex toolbox functions. Implementing the Holoscan pipeline with MATLAB requires four steps: Creating a MATLAB function, using the GPU Coder? Generate accelerated CUDA code, create a Holoscan operator wrapper, and recompile the Holoscan application with the new MATLAB operator. This makes it possible to create a fully functional software-defined workflow for medical devices. The additional verification and validation capabilities integrated with MATLAB and Holoscan enable the development of software-defined workflows in compliance with industry regulations and standards, including IEC 62304.